A grand challenge from the wind energy industry is to provide reliable forecasts on mountain winds several hours in advance at microscale (∼100 m) resolution. This requires better microscale wind-energy physics included in forecasting tools, for which field observations are imperative. While mesoscale (∼1 km) measurements abound, microscale processes are not monitored in practice nor do plentiful measurements exist at this scale. After a decade of preparation, a group of European and U.S. collaborators conducted a field campaign during 1 May–15 June 2017 in Vale Cobrão in central Portugal to delve into microscale processes in complex terrain. This valley is nestled within a parallel double ridge near the town of Perdigão with dominant wind climatology normal to the ridges, offering a nominally simple yet natural setting for fundamental studies. The dense instrument ensemble deployed covered a ∼4 km × 4 km swath horizontally and ∼10 km vertically, with measurement resolutions of tens of meters and seconds. Meteorological data were collected continuously, capturing multiscale flow interactions from synoptic to microscales, diurnal variability, thermal circulation, turbine wake and acoustics, waves, and turbulence. Particularly noteworthy are the extensiveness of the instrument array, space–time scales covered, use of leading-edge multiple-lidar technology alongside conventional tower and remote sensors, fruitful cross-Atlantic partnership, and adaptive management of the campaign. Preliminary data analysis uncovered interesting new phenomena. All data are being archived for public use.

The New European Wind Atlas project will create a freely accessible wind atlas covering Europe and Turkey, develop the model chain to create the atlas and perform a series of experiments on flow in many different kinds of complex terrain to validate the models. This paper describes the experiments of which some are nearly completed while others are in the planning stage. All experiments focus on the flow properties that are relevant for wind turbines, so the main focus is the mean flow and the turbulence at heights between 40 and 300 m. Also extreme winds, wind shear and veer, and diurnal and seasonal variations of the wind are of interest. Common to all the experiments is the use of Doppler lidar systems to supplement and in some cases replace completely meteorological towers. Many of the lidars will be equipped with scan heads that will allow for arbitrary scan patterns by several synchronized systems. Two pilot experiments, one in Portugal and one in Germany, show the value of using multiple synchronized, scanning lidar, both in terms of the accuracy of the measurements and the atmospheric physical processes that can be studied. The experimental data will be used for validation of atmospheric flow models and will by the end of the project be freely available. This article is part of the themed issue ‘Wind energy in complex terrains’.

We present results of the GABLS3 model intercomparison benchmark revisited for wind energy applications. The case consists of a diurnal cycle, measured at the 200-m tall Cabauw tower in the Netherlands, including a nocturnal low-level jet. The benchmark includes a sensitivity analysis of WRF simulations using two input meteorological databases and five planetary boundary-layer schemes. A reference set of mesoscale tendencies is used to drive microscale simulations using RANS k- and LES turbulence models. The validation is based on rotor-based quantities of interest. Cycle-integrated mean absolute errors are used to quantify model performance. The results of the benchmark are used to discuss input uncertainties from mesoscale modelling, different meso-micro coupling strategies (online vs offline) and consistency between RANS and LES codes when dealing with boundary-layer mean flow quantities. Overall, all the microscale simulations produce a consistent coupling with mesoscale forcings.

This study addresses the importance of enhancing traditional fluid-flow solvers by introducing a Machine Learning procedure to model pressure fields computed by standard fluid-flow solvers. The conventional approach involves enforcing pressure–velocity coupling through a Poisson equation, combining the Navier–Stokes and continuity equations. The solution to this Poisson equation constitutes a substantial percentage of the overall computational cost in fluid flow simulations, therefore improving its efficiency can yield significant gains in computational speed. The study aims to create a versatile method applicable to any geometry, ultimately providing a more efficient alternative to the conventional pressure solver. Machine Learning models were trained with flow fields generated by a Computational Fluid Dynamics solver applied to the confined flow over multiple geometries, namely wall-bounded cylinders with circular, rectangular, triangular, and plate cross-sections. To achieve applicability to any geometry, a method was developed to estimate pressure fields in fixed-shape blocks sampled from the flow domain and subsequently assemble them to reconstruct the entire physical domain. The model relies on multilayer perceptron neural networks combined with Principal Component Analysis transformations. The developed Machine Learning models achieved acceptable accuracy with errors of around 3%. Furthermore, the model demonstrated enhanced computational efficiency, outperforming the classical PISO algorithm by up to 30 times.